If America Leads China by 23x in AI Spending, Why Fear Kimi K3?

America spends 23 times more than China on private AI. Yet one Chinese model, Kimi K3, has Washington talking about bans again.

Stanford’s 2026 AI Index Report counts $285.9 billion in US private AI investment for 2025. China recorded just $12.4 billion. Even so, Commerce is not currently moving toward a ban, a person familiar with the matter says.

Why Kimi K3 Has Washington Worried

Kimi K3 comes from Chinese startup Moonshot AI. It is open-weight, so anyone can download it and run it on their own servers. Kimi K3’s rapid rise in coding tests rattled US chip stocks last week.

That success revived old plans. According to Axios, the Commerce Department last year weighed putting Chinese AI labs on its Entity List. That is the same trade blacklist that hit Huawei in 2019.

There was more. The National Security Agency (NSA) considered a public warning about Chinese AI. The White House explored making US firms legally liable if a hosted Chinese model got breached.

Officials focused on innovation killed those ideas. However, the security hawks are louder now, and Kimi K3 handed them fresh ammunition.

“We are at a critical inflection point in AI policy. The leading closed labs, already a duopoly in terms of AI model revenue, want the government to eliminate their open-source competition,” David Sacks, an outside White House AI adviser, indicated.

Meanwhile, demand for the model keeps growing. Moonshot paused new subscriptions within 48 hours of launch. It is also preparing a Hong Kong IPO.

What the 23x AI Spending Gap Hides

The 23x figure is an interesting outlook, with Stanford’s own report saying the comparison hides part of China’s spending. State-run guidance funds pumped an estimated $184 billion into Chinese AI firms between 2000 and 2023.

The US number has its own catch. Most of it flows to a handful of giants. Funding rounds above $1 billion nearly doubled to 28 in 2025, led by OpenAI’s $40 billion raise.

Price explains why US firms keep choosing Chinese models anyway. DeepSeek’s V4 Pro charges $0.87 per million output tokens. Anthropic’s Claude Fable 5 lists at $50 for the same amount.

The savings are real. Coinbase CEO Brian Armstrong said in June that the exchange runs GLM 5.2 and Kimi K2.7 Code, cutting its AI bill roughly in half.

Washington has felt this shock before. DeepSeek’s rise erased a record $589 billion from Nvidia’s value in one day in January 2025. The ban proposals now back on the table first circulated in the months after that crash.

A ban may not even work. Kimi K3’s weights already sit on public repositories, and Tom’s Hardware reported that recalling them is nearly impossible. Jack Dorsey echoed an open source AI warning that restrictions could raise costs, while Alibaba’s Qwen3.8-Max debut shows more challengers are coming either way.

The coming weeks will show whether Washington leans on warnings and procurement rules or lets the market decide. The scoreboard says America is far ahead. Kimi K3, like DeepSeek before it, suggests the race may come down to capability, not capital.

The post If America Leads China by 23x in AI Spending, Why Fear Kimi K3? appeared first on BeInCrypto.